Siemens RUGGEDCOM ROX II User Manual page 679

RUGGEDCOM ROX II
CLI User Guide
• Improves performance for large ring network topologies (up to 160 switches)
eRSTP is also compatible with standard RSTP for interoperability with commercial switches.
For example, in a network comprised of 15 RUGGEDCOM hardened Ethernet switches in a ring topology, the
expected fault recovery time would be less than 75 ms (i.e. 5 ms x 15). However, with eRSTP, the worst case fault
recovery time is less than 26 ms.
Section 14.3.1.7
Fast Root Failover
Siemens's Fast Root Failover feature is an enhancement to RSTP that may be enabled or disabled. Fast Root
Failover improves upon RSTP's handling of root bridge failures in mesh-connected networks, resulting in slightly
increased failover times for some non-root bridge scenarios.
IMPORTANT!
In networks mixing RUGGEDCOM and non-RUGGEDCOM switches, or in those mixing Fast Root Failover
algorithms, RSTP Fast Root Failover will not function properly and root bridge failure will result in an
unpredictable failover time. To avoid potential issues, note the following:
• When using the Robust algorithm, all switches must be RUGGEDCOM switches
• When using the Relaxed algorithm, all switches must be RUGGEDCOM switches, with the exception
of the root switch
• All RUGGEDCOM switches in the network must use the same Fast Root Failover algorithm
Two Fast Root Failover algorithms are available:
• Robust – Guarantees a deterministic root failover time, but requires support from all switches in the network,
including the root switch
• Relaxed – Ensures a deterministic root failover time in most network configurations, but allows the use of a
standard bridge in the root role
NOTE
The minimum interval for root failures is one second. Multiple, near simultaneous root failures (within
less than one second of each other) are not supported by Fast Root Failover.
Fast Root Failover and RSTP Performance
• Running RSTP with Fast Root Failover disabled has no impact on RSTP performance.
• Fast Root Failover has no effect on RSTP performance in the case of failures that do not involve the root bridge
or one of its links.
• The extra processing introduced by Fast Root Failover significantly decreases the worst-case failover time
in mesh networks, with a modest increase in the best-case failover time. The effect on failover time in ring-
connected networks, however, is only to increase it.
Recommendations On the Use of Fast Root Failover
• It is not recommended to enable Fast Root Failover in single ring network topologies
• It is strongly recommended to always connect the root bridge to each of its neighbor bridges using more than
one link
Fast Root Failover
Chapter 14
Network Redundancy
633